A Bispecific, Tetravalent Antibody Targeting Inflammatory and Pruritogenic Pathways in Atopic Dermatitis

Abstract

Inhibition of IL-4/IL-13 signaling has dramatically improved the treatment of atopic dermatitis (AD). However, in many patients, clinical responses are slow to develop and remain modest. Indeed, some symptoms of AD are dependent on IL-31, which is only partially reduced by IL-4/IL-13 inhibition. Thus, there is an unmet need for AD treatments that concomitantly block IL-4/IL-13 and IL-31 pathways. We engineered NM26-2198, a bispecific tetravalent antibody designed to accomplish this task. In reporter cell lines, NM26-2198 concomitantly inhibited IL-4/IL-13 and IL-31 signaling with a potency comparable with that of the combination of an anti-IL-4Rα antibody (dupilumab) and an anti-IL-31 antibody (BMS-981164). In human PBMCs, NM26-2198 inhibited IL-4-induced upregulation of CD23, demonstrating functional binding to FcγRII (CD32). NM26-2198 also inhibited the secretion of the AD biomarker thymus and activation-regulated chemokine (TARC) in blood samples from healthy human donors. In male cynomolgus monkeys, NM26-2198 exhibited favorable pharmacokinetics and significantly inhibited IL-31-induced scratching at a dose of 30 mg/kg. In a repeat-dose, good laboratory practice toxicology study in cynomolgus monkeys, no adverse effects of NM26-2198 were observed at a weekly dose of 125 mg/kg. Together, these results justify the clinical investigation of NM26-2198 as a treatment for moderate-to-severe AD.

Keywords: Atopic dermatitis; Cytokines; Inflammatory skin diseases; Pruritus.

Figure 1

RNA-seq analysis of healthy human skin biopsies from 3 donors treated with IL-4, IL-13, and IL-31 at 50 ng/ml for 72 h.(a) Volcano plots showing genes whose expression was significantly altered (red) by IL-4 + IL-13 (left), IL-31 (middle), and IL-4 + IL-13 + IL-31 (right) treatments. (b) Heatmap showing relative expression of specific genes according to treatment condition and donor based on row Z-scores of all samples in the study, of which FLG, FLG2,IL-24,CXCL-11, IL13Ra2,IL-33,LCN2, K6B, GZMB, CCL17, and CCL13 are known to play an important role in AD. Other important genes known to play a role in inflammatory and neural processes, keratinocytes, and/or barrier function are also highlighted on the right. (c) Bar plot showing changes in expression (log-transformed adjusted P-values) of gene clusters associated with selected gene ontology biological processes terms after treatment with the indicated ILs. There was no enrichment of the shown terms for IL-31. (d, e) Bar plots showing top enriched gene ontology biological processes terms after treatment with cytokines. AD, atopic dermatitis; h, hour; K6B, keratin 6B; RNA-seq, RNA sequencing.

Figure 2

Schematic depiction (left) and structural model (right) of NM26-2198. The structural model was prepared using BIOVIA Discovery Studio software.

Figure 3

NM26-2198 binding kinetics and inhibition of cell signaling.(a) SPR sensorgrams of NM26-2198 binding to IL-4Ra in multicycle kinetic mode. (b) SPR sensorgram of NM26-2198 binding to IL-31 in single-cycle kinetic mode. Colored lines represent the experimental data, and black lines show the fit of a 1:1 interaction model. (c) SPR sensorgram showing concomitant binding of NM26-2198 to human IL-4R (immobilized) and IL-31. (d) SEAP concentration in the supernatant of HEK-Blue IL-4/IL-13 STAT6 reporter gene cells measured using the QUANTI-Blue colorimetric assay and reported as mean (SD) optical density. Cells were cultured in the presence of IL-4 (0.05 ng/ml, 23.8 pM) for 24 hours to induce SEAP secretion. Culture media also contained the indicated concentrations of dupilumab or NM26-2198 in the presence or absence of excess IL-31 (9.9 nM). (e) Mean (SD) SEAP concentrations in similar experiments as d except that cells were cultured in the presence of IL-13 (0.3 ng/ml, 23.8 pM) instead of IL-4. (f) Mean (SD) chemiluminescence of engineered U2OS cells indicating IL-31–induced heterodimerization of IL-31RA and OSMR after 6 hours incubation with IL-31 (10 ng/ml, 0.63 nM) and the indicated concentrations of BMS-981164 or NM26-2198 in the presence or absence of excess IL-4R extracellular domain (50 nM). (g) Mean (SD) CCL2 concentrations in the supernatant of cultured BEAS-2B cells after 24-hour incubation with the indicated ILs and antagonists. Antagonist concentrations were 1 mg/ml (5 nM dupilumab and 5 nM BMS-981164, 6.7 nM NM26-2198). Results are mean (SD) of 1 independent experiment. P-values are based on Dunnett’s test performed after an overall P = .012 by ANOVA. ∗ P < .05 and ∗∗ P < .01. (h) CCL2 concentrations in supernatant of BEAS-2B cells after 24-hour incubation with IL-4 (1 ng/ml) and IL-31 (10 ng/ml) and the indicated concentrations of dupilumab + BMS-981164 or NM26-2198. a and b are representative of 3 experiments, and c–h correspond to 1 experiment. Duplicate wells were used for experiments shown in d–h. OSMR, oncostatin M receptor; SEAP, secreted embryonic alkaline phosphatase; SPR, surface plasmon resonance; STAT6, signal transducer and activator of transcription 6.

Figure 4

NM26-2198 inhibition of IL-4–induced CD23 expression in human monocytes and B cells and of IL-4–induced TARC secretion in whole human blood.(a–c) Geometric mean CD23 expression in (a) monocytes, (b) naïve B cells, or (c) memory B cells measured by flow cytometry after 48-hour incubation of PBMCs with IL-4 (2 ng/ml) and the indicated concentrations of NM26-2198, its IgG1-Fc silenced version PRO2142, or dupilumab. (d–f) Secretion of TARC (CCL17) in specimens of whole human blood from 3 donors after 24-hour incubation with IL-4 (1 ng/ml) and the indicated concentrations of dupilumab or NM26-2198. Secretion was quantified by ELISA, and the values shown are means (SD) of duplicate measurements at each concentration. TARC, thymus and activation-regulated chemokine.

Figure 5

NM26-2198 pharmacokinetics and inhibition of IL-31–induced scratching in cynomolgus monkeys.(a, b) Mean (SD) serum drug levels at indicated times after intravenous (green) or subcutaneous (orange) administration of single doses of NM26-2198 to cynomolgus monkeys (n = 6 per treatment condition, except n = 5 for 3 mg/kg group). (c) Treatment timeline of scratching study as described in the text. Monkeys were grouped according to baseline treatment (day −28 to day −18). (d) Number of scratching events during 3 hours after IL-31 injection minus 3× the number of scratching events during the 1 hour preceding IL-31 injection. Filled circles indicate results for each monkey; bars indicate the mean for each treatment condition. ∗P < .05 (0.017) for monkeys treated with 30 mg/kg versus vehicle on the basis of Dunnett’s test performed after an overall P = .003 by ANOVA.